Is the yield curve still a good predictor of real economic activity?

Abstract

An extensive body of empirical research documents that the slope of the yield curve - the spread between long- and short-term interest rates - is a powerful tool for forecasting economic growth. This study investigates whether the yield curve is still a good predictor of real economic activity and thus examines if this long-standing empirical regularity still holds, particularly in the light of the turbulence brought about by the recent global financial crisis. Towards this end, a series of forecasting equations are employed, which predict real GDP growth using the current yield spread and current GDP growth, based on an extended U.S. dataset stretching from 1961 to 2011. The spread used is that between the 10-year and the three-month rate, as is customary in the literature. For the long-term interest rate, constant-maturity treasuries (CMTs) data is used to avoid the problem of mismatching maturities. The forecasting performance of the equations is gauged along two key dimensions by assessing how well they perform (i) out-of-sample, and (ii) compared to a standard benchmark, specifically, the naïve random walk model. The results are subjected to a series of robustness checks not previously performed in the literature. Despite data limitations, the study is extended to include two other countries; France, the only other country for which CMT data is available, and Malta. The results reveal that there is a relationship between the current yield spread and future GDP growth in the case of U.S. and France. However the relationship does not hold for Malta. When subjected to an out-of-sample forecasting exercise, the forecasting equation for France using the spread performed substantially better than the naïve random walk model. For the U.S., the forecast performance of the equation was worse than that of a naïve random walk model. However, the difference in the predictive ability of the two forecasting techniques is minimal. Thus this shows that the yield curve can return to being one of the best predictor of GDP growth in the near future as it is gaining ground when compared to other forecasting techniques.